From Simple Refill Calls to Clinical-Grade Pharmacy Conversations

How Hospitals Can Future-Proof Pharmacy Communications with AI

Hospital pharmacy teams are under growing pressure. Rising patient volumes, medication complexity, and staff shortages are turning routine outbound calls into a major operational burden.

The operational impact of manual communications is no longer just a feeling; it is a measurable clinical risk. Supporting data shows that hospital pharmacists now spend almost a quarter (25%) of their work time dealing with interruptions, the majority of which are phone calls. These disruptions occur with a staggering frequency of 5 to 13 times per hour, creating a fragmented workflow where interruptions contribute to over 50% of incidents involving medication errors.¹

Over the past few months, we’ve spent significant time speaking with hospital pharmacy leaders across different health systems. While their organizations vary, the challenges they describe are strikingly consistent.

In this post, we want to summarize what we’ve learned from those conversations, and share the patterns we see emerging as hospitals explore AI to support pharmacy communications.

Recurring themes highlighted by pharmacy leaders include:

• Pharmacists and technicians spend too much time on repetitive calls
• Difficulty scaling pharmacy outreach without compromising patient safety
• Frustration with tools built for call-center automation or billing, not for healthcare
• Generic EHR outbound modules or contact-center solutions often work for basic campaigns but fail when conversations get complex

Faced with these operational and clinical pressures, pharmacy leaders repeatedly ask the same question.

“Can AI help automate pharmacy calls safely?”

‍

The reality of pharmacy calls in hospitals

In practice, pharmacy communications are bifurcated into two distinct categories, each requiring a different level of intelligence and oversight.

1. Simple, high-volume calls

These calls are predominantly repetitive and administrative, making them the primary candidates for automation. Key examples include:

• Refill confirmation
• Medication Readiness or pickup reminders

However, even a “simple” refill conversation can quickly become more complex.

In practice, during what starts as a basic refill confirmation, patients may:

• Mention a recent change of address or living situation
• Say they have stopped taking the medication
• Express confusion about dosing or timing
• Report side effects or adverse reactions

When tools are not built for healthcare, they often fail to detect and appropriately handle these moments. As a result, two risks emerge:

• Clinical risk, if important safety signals go unnoticed
• Poor patient experience, when patients feel unheard or unable to escalate

The ability to recognize and understand these situations, and to manage them safely, is critical, even when the original intent of the call was simple.

2. Complex, safety-critical conversations

These interactions are not mere "outreach"; they are clinical consultations that require deep context and triage.

Pharmacy teams frequently encounter:

• Patients who are not adherent
• Questions about dosing or timing
• Side effects or adverse reactions
• Confusion about medication changes after discharge

These are not simple call-center conversations.

They require clinical context, careful triage, and reliable escalation pathways. Treating them as generic outreach inevitably increases risk and frustration for patients.

To make things even harder, some of the calls will be inbound, making it impossible to discern beforehand if the patient is calling for a simple request or if it's going to need a more sophisticated conversation.

The true value of a clinical AI agent isn't just in the calls it automates, but in the intelligence it captures. By identifying clinical flags, even during a routine refill, the system ensures professional oversight: it can autonomously resolve low-risk clinical doubts using hospital protocols, while immediately notifying pharmacists of any high-stakes risk.

This allows the team to focus on where their expertise is vital, while the AI manages both the administrative volume and basic clinical guidance.

‍

Addressing a common concern: Do patients trust and engage with AI clinical calls?

In many hospitals, some care team members are understandably skeptical that an AI can deliver the same level of patient satisfaction, empathy, or service quality as a human.

This concern comes up frequently in our conversations with pharmacy and clinical leaders, and it’s an important one to address openly.

What we see in practice is that confidence builds through experience. As teams observe how patients actually respond, and as pharmacists experience the impact on their daily workload, perceptions begin to change.

Across deployments, hospitals consistently observe:

• High patient engagement and call completion rates
• Strong patient satisfaction with AI-led calls
• Fewer unanswered or avoided outreach attempts

At the same time, care teams begin to notice tangible benefits:

• Less time spent on repetitive, low-value calls
• Fewer interruptions during day-to-day work
• More focused involvement, only when expertise is truly needed

From the patient's perspective, feedback often highlights:

• Timely, proactive communication
• Clear and structured conversations
• The ability to speak at their own pace, without feeling rushed

Over time, these shared observations, from both patients and care teams, tend to shift the conversation from skepticism to trust.

‍

Why phased, guided implementation makes the difference

Trust in AI is rarely built all at once, and it shouldn’t be.

In our experience, the most successful teams adopt AI through a phased and carefully guided approach that deliberately balances speed, safety, and technical risk.

Early phases often focus on simple, low-risk conversations and lightweight data exchange, without requiring deep integration into legacy pharmacy or EHR systems from day one. This reduces implementation friction while still allowing teams to move quickly.

Critically, results are made visible from the start.

Structured call outcomes, patient responses, and escalation signals are fed into dashboards and BI tools, giving clinical, operational, and leadership stakeholders clear visibility into performance, safety signals, and impact.

Successful teams typically:

1. Start with simple, low-risk conversations
2. Share results transparently with the care team
3. Involve pharmacists in reviewing conversations and outcomes
4. Expand gradually as confidence and trust grow

This approach consistently helps:

• Demonstrate real impact early
• Build buy-in from more hesitant team members
• Align clinical, operational, and leadership stakeholders

Over time, what begins as cautious experimentation often becomes strong advocacy, driven by evidence.

‍

One system, multiple levels of complexity: The power of a unified platform

At Tucuvi, we designed LOLA, our AI clinical agent, specifically for healthcare conversations, including hospital pharmacy workflows.

This allows hospitals to:

• Phase 1: Start with Simple Use Cases: Automate high-volume tasks such as refill confirmations, medication availability checks, pickup confirmations, and return-to-stock prevention calls.  
• Phase 2: Grow into Advanced Use Cases: Transition naturally into complex interactions, including medication adherence assessments, high-risk medication follow-up, detection of clinical red flags, or Assisted Medication Reconciliation.

All on the same platform, with the same AI Agent, ensuring consistency and the same quality and experience for both patients and care teams.

Quality and safety: Human-in-the-Loop by design

In pharmacy, automation without oversight is not an option. That’s why Human-in-the-Loop is a foundational part of our approach:

• Supervised Improvement: Conversations are monitored and continuously improved based on real-world interactions.  
• Protocol-Driven Escalation: Clinical protocols guide all decision-making, ensuring the AI recognizes when a situation requires professional judgment.  
• Focused Expertise: Pharmacists intervene only when their specialized expertise is truly required, maintaining a perfect balance between autonomy and oversight.

This balance between autonomy and oversight is critical to building long-term trust.

Beyond efficiency: reducing workload and improving care

When pharmacy calls are handled by a clinical AI rather than a generic call-center tool:

• Reclaimed Clinical Time: Care teams can shift their focus back to higher-value clinical work.  
• Enhanced Patient Satisfaction: Patients receive consistent, empathetic, and proactive communication.  
• Risk Detection & Effective Escalation: Safety signals and clinical red flags are captured reliably, so no escalation falls through the cracks.
• Actionable Data: AI documents 100% of calls with structured results. Hospitals gain structured clinical insights rather than just basic call outcomes, directly improving the quality of care.

Efficiency improves, but so does quality of care.

‍

Looking ahead: The Strategic Evolution of Pharmacy Automation

Hospitals that succeed with AI in pharmacy don’t think in terms of isolated campaigns. They think in terms of long-term capability.

They start simple. They build trust. And when they are ready, they expand into more sophisticated clinical conversations, without changing vendors, systems, or patient experience.

That’s how AI becomes a natural extension of the pharmacy team, not a replacement, and not a call-center workaround.

The question is no longer whether AI can support pharmacy communications. The real challenge for modern health systems is implementing it in a way that proves value, ensures safety, and evolves alongside the complex needs of their patients

Modernizing pharmacy communications is a strategic step toward reducing medication errors and clinical burnout. This approach ensures that clinical expertise is focused on high-stakes interventions while routine logistics are managed with consistent, protocol-driven precision.

Schedule a demo to see how LOLA can be integrated into your pharmacy’s clinical communication standards.

‍

References

1. Karia A, Norman R, Robinson S, Lehnbom E, Laba TL, Durakovic I, Balane C, Joshi R, Webster R. Pharmacist's time spent: Space for Pharmacy-based Interventions and Consultation TimE (SPICE)-an observational time and motion study. BMJ Open. 2022 Mar 2;12(3):e055597. doi: 10.1136/bmjopen-2021-055597. PMID: 35236731; PMCID: PMC8896034.

‍